Get to know Sahaana Suri, Member of Technical Staff; AI at PlayerZero.
Sahaana leads AI research focused on making PlayerZero sharper with every interaction. In practice, that means turning the messy, implicit knowledge engineers carry in their heads into something structured, testable, and durable — our engineering world model that preserves intent even as code and systems evolve. She’s also building the team and culture needed to push that frontier forward (we’re hiring 👋).
Before joining PlayerZero, Sahaana earned her PhD in AI/ML systems at Stanford, where she focused on data curation: how to decide what matters, what doesn’t, and how to build systems that improve over time. Along the way, she’s worked at Google, interned at Amazon and Twitter, and even started her own company. She brings these experiences to PlayerZero by bridging theory and practice, ensuring research translates into practical and scalable systems that are useful in real-world engineering environments.
Sahaana had been tracking PlayerZero long before she officially joined. Back when it was still a summer project built by Animesh Koratana and Maria Vinokurskaya, she saw early signals that it was onto a completely new way of thinking about the gap between intended behavior and runtime reality. Over time, as the product and team matured, that conviction only grew, until joining felt inevitable.
What ultimately pulled her in was the combination of a product she genuinely wanted to use herself, clear signals of real-world traction, and a small, high-trust team moving with urgency. It felt like the right place to build, at the right time.
What excites her most is the core problem of taking intuition — patterns senior engineers recognize instantly — and turning it into systems that can reason, generalize, and improve. Recently, she built a system that learns from how experienced support engineers decide when to escalate issues. Instead of relying on static rules, it captures subtle patterns (e.g., error combinations, customer context, historical resolution paths) and turns them into something operational. The result: teams have seen unnecessary escalations drop by 60% without missing critical issues.
For Sahaana, that’s the crux of the work. Not just building models, but building systems that understand how engineers think and get better over time. “Figuring out how to represent tribal knowledge in a durable, testable way is the hardest part,” she says, “and also the most rewarding.”
Outside of work, she’s usually at the gym. If you ask what animal she’d be, the answer depends on the day — sometimes a cheetah, sometimes a sea slug. And if forced to pick a single fruit for life, she’s still split between passion fruit (flavor) and bananas (practicality).